Large Genome Analysis Identifies Three Subtypes of Prostate Cancer

Pam Harrison

September 27, 2016

BOSTON — Researchers have identified and rigorously validated three intrinsic molecular subtypes of primary prostate cancer that independently predict a patient's risk of developing distant metastases after undergoing surgical removal of the prostate, new research shows.

The same three subtypes also appear to be able to predict who will respond most favourably to postoperative radiation therapy after radical prostatectomy.

"This is exactly the type of research that we truly need and I can't wait to see how we can incorporate this kind of information into the mainstream and into all of our ROGT [Radiation Therapy Oncology Group] trials," press session moderator, Colleen Lawton, MD, Medical College of Wisconsin, Milwaukee, told a briefing here.

"This is really tapping into something we haven't had before," she said.

The study was presented here during the American Society for Radiation Oncology (ASTRO) 2016 Annual Meetingin Boston, Massachusetts.

Nine Large Cohorts

The new research was presented by Daniel Spratt, MD, assistant professor of radiation oncology, University of Michigan, Ann Harbor. The research team assembled nine large cohorts, two of them prospective cohorts, from which they had over 4200 samples taken from men who had undergone radical prostatectomy for localized prostate cancer.

"This makes this analysis the largest genomic analysis ever in primary prostate cancer," Dr Spratt told journalists.

Investigators then "fed" the genomic data into an unbiased clustering algorithm and told it to identify the 100 most highly and variably expressed genes to generate a heat map.

From this map, "we identified threesubtypes, which we termed A, B, and C," Dr Spratt noted.

The researchers then validated that the three subtypes were highly reproducible across all of the cohorts, and they looked for what this biology might mean for patients. They also sought to determine the potential prognostic impact that each of these three intrinsic subtypes might have on clinical outcomes.

"We found that patients with subtype A had the most favorable prognosis," Dr Spratt reported.

Furthermore, after correcting for confounding variables, such as Gleason score and prostate-specific antigen levels, "not only do subtype A patients retain the most favorable prognosis but we found that molecular subtypes independently predict a patient's risk of distant metastases," he added.

Specifically, distant metastasis-free survival rates at 10 years were 73.6% for men with subtype A, 64.4% for men with subtype B, and 57.1% for men with subtype C.

The team next asked whether the same subtypes might be able to predict which men would benefit the most from postoperative radiation therapy.

To answer this question, they analyzed a sample of 919 men who did not receive postoperative radiation therapy after radical prostatectomy and found that, again, men with subtype A had a much more favorable prognosis than those with subtypes B and C.

The researchers then compared an analysis of 350 men who received postoperative radiation therapy after radical prostatectomy and found no difference in outcomes between men with subtype A and those with subtype B or C.

However, by quantifying the magnitude of benefit from radiation therapy across the three subtypes, "we found that patients with subtypes B and C had a greater than two-fold improved response to postoperative radiation therapy than patients with subtype A," Dr Spratt noted.

"And this suggests that this may be the first predictive biomarker to tell us which patients are benefitting the most from radiation therapy," he added.

Clinical Implications

As Dr Spratt speculated, once physicians know a prostate tumor's genetic makeup, ideally patients could be randomly assigned into different clinical trial groups and, provided the treatment groups are balanced with respect to the different subtype groups represented, they will get a truer picture of the benefit of a given intervention.

"Genomic risk also tells us that a patient we think is low risk clinically may actually have more aggressive disease, so they may be high risk genomically," Dr Spratt said.

Conversely, physicians may have patients who are considered high risk on the basis of clinical parameters alone but genomically are destined to behavior much more like a low-risk patient.

"Certainly, we are likely to find subsets of patients who we are going to treat differently because we've learned that they respond better to radiation therapy or hormonal treatment or some sort of combination," Dr Lawton commented, after hearing the presentation.

"This is the kind of information we don't have right now but which would be exceedingly helpful," she added.

American Society for Radiation Oncology (ASTRO) 2016 Annual Meeting. Abstract 6. Presented September 25, 2016.

Follow Medscape Oncology on Twitter: @MedscapeOnc


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: